Project Summary/Abstract Eukaryotic cells display remarkable metabolic plasticity to maintain steady-state flow of energy and metabolites in constantly changing cellular environments. This plasticity is particularly relevant in the context of cancer as cancer cells require substantial metabolic alteration to live in their nutrient and oxygen poor microenvironment. Understanding how cancer cells rewire their metabolism in tumors is one of the central challenges in cancer metabolism field. However, studying tumor metabolism in vivo adds many layers of biological complexity. First of all, the nutrient composition of most human tumor types remains to be defined, and it is still not clear which components are abundant or limited, or to what extent they vary between malignancies. Another major challenge is that tumor microenvironment is extraordinarily complex, containing a diversity of intermingled non-cancerous cell types, which provide metabolic support for cancer cells. In the last decade, metabolite profiling of whole tumors had a major impact on cancer metabolism field by enabling simultaneous analysis of hundreds of metabolites. These approaches, while extremely valuable in providing a broad view of entire biochemical pathways, do not scale to the cellular and metabolic diversity of tumors. To overcome this challenge, we need a conceptually new and robust method to monitor the metabolic status of each cell type under different metabolic stress conditions in vivo. To address this, my lab will develop technologies that enable the use of metabolomics to map key metabolites within discrete cell types and environmental conditions of tumors. These technologies are based on measuring metabolites in captured cellular compartments to determine precise cellular metabolic identity so that we can use metabolite signatures as entry points to exploit metabolic features for cancer therapy. In this essay, I will focus on mitochondria, a central biosynthetic hub for cellular metabolism, and describe how we will use mitochondrial metabolomics to study tumor metabolic heterogeneity in vivo. I will first apply this method to map metabolites of cancer and stromal cells in pancreatic tumors, whose hallmarks include dramatic alterations in nutrient availability and utilization, and in tumor-stroma communication. I will then generate a new set of tools to engineer mitochondria so that we can capture metabolites of cells under metabolic stress conditions observed in pancreatic tumors (i.e hypoxia, energy stress). Finally, I will exploit these transformative tools to elucidate the metabolic adaptations associated with therapeutic response and resistance using KRAS-driven models of pancreatic cancer. Overall, these approaches will provide unprecedented opportunities to delineate tumor metabolic heterogeneity.